Medication-induced diabetes during inductiontreatment for ALL, an early marker for futuremetabolic risk?
Yeshayahu Y, Koltin D, Hamilton J, Nathan PC, Urbach S.Medication-induced diabetes during induction treatment for ALL, an earlymarker for future metabolic risk?.Pediatric Diabetes 2014.
Medication-induced diabetes (MID) is seen in children treated for acutelymphoblastic leukemia (ALL) mostly during induction, due to the use ofl-asparaginase and glucocorticoids. Our objective was to assess whether MIDduring induction, is a risk factor for future impaired glucose tolerance (IGT),diabetes, or metabolic syndrome. Ninety survivors of pediatric ALL, ages10 yr and older were recruited, 30 with history of MID and 60 controls.Waist/height ratio >0.5 was considered as an increased risk for centraladiposity and insulin resistance. Lipid profile and an oral glucose tolerancetest (OGTT) were performed. Study patients were older than controls (17.2 vs.14.9, p < 0.05). The groups had similar sex distribution, body mass index(BMI) z-score, and Tanner staging. A waist/height ratio of >0.5 was seen in 60and 31.7% of the study and control groups, respectively (p = 0.01). Increasedfrequency of IGT in the study group compared with the control group wasseen (13.3 and 1%, respectively) (p = 0.07). We observed a trend toward higherproportion of patients with multiple features of metabolic syndrome in thestudy compared with control group (16.7 vs. 5%, p = 0.09). In conclusion,MID during induction may be an early marker for metabolic disturbanceslater in life. The higher rates of increased waist/height ratio, and subjects withmultiple metabolic syndrome features, may predict a metabolic risk inchildren with history of MID. Rates of IGT were four fold higher in the studygroup although not statistically significant. MID may be a ‘red flag’ indicatingthe need for ongoing metabolic screening and lifestyle modifications toprevent future metabolic disease.
Yonatan Yeshayahua,b,†,Dror Koltinc,d,†,Jill Hamiltonc,d, Paul C.Nathand,e and StaceyUrbachc,d
aDivision of Endocrinology, SafraChildren’s Hospital, Ramat Gan, Israel;bDepartment of Pediatrics, Tel-AvivUniversity, Tel-Aviv, Israel; cDivision ofEndocrinology, The Hospital for SickChildren, Toronto, ON, Canada;dDepartment of Pediatrics, Universityof Toronto, Toronto, ON, Canada; andeDivision of Haematology Oncology,The Hospital for Sick Children,Toronto, ON, Canada†Contributed equally to the study andthe manuscript.
Corresponding author: YonatanYeshayahu, MD, MHA,Division of Endocrinology,Safra Children’s Hospital,Tel-Hashomer,Ramat Gan 52621,Israel.Tel: (972) 3-5305015;fax: (972) 3-5305055;e-mail:[email protected]
Submitted 26 November 2013.Accepted for publication 21February 2014
Acute lymphoblastic leukemia (ALL) is the mostcommon childhood malignancy accounting for onethird of cases of childhood cancer (1). Medication-induced diabetes (MID) has been well describedin children receiving therapy for ALL (2). This
complication is seen mostly during remission induc-tion, and is attributed to specific chemotherapeuticagents, such as l-asparaginase causing pancreaticβ-cell dysfunction, and glucocorticoids leading toinsulin resistance (2, 3). The prevalence of MID during
1
Yeshayahu et al.
ALL therapy in childhood has been reported to bebetween 9.7 and 20.4% (2–5). While this complicationis quite common and self resolving in most cases, it isunknown whether its occurrence carries any long-termmetabolic implications. Several studies have evalu-ated the association between MID during inductionand long-term outcome from a survival perspective,with inconsistent results (6–8). None of these studieshave examined whether MID during ALL inductionportends future metabolic risk.
With current treatment protocols the 5 yr survivalrates in ALL exceed 85% (9). Survivors are atconsiderable risk of long-term complications resultingfrom their disease and its therapy, including earlycardiovascular morbidity and mortality secondary totreatment with anthracyclines. Death from cardiac-related events was shown to be 8.2 times more likely incancer survivors compared with the general population(10), and adult cancer survivors have been shown tohave a higher prevalence of features of the metabolicsyndrome (11). This predisposition for the metabolicsyndrome further increases the cardiovascular riskexperienced by survivors of ALL. The mechanismsand risk factors leading to the development of featuresof the metabolic syndrome are not well understood. Asingle study in pediatric patients confirmed impairedpancreatic β-cell function in ALL survivors, but thisstudy did not assess the relationship with previousdevelopment of MID during treatment (12) .
The aim of our study was to assess whether thedevelopment of MID during remission induction, inpediatric patients with ALL, is a risk factor fordeveloping impaired glucose tolerance (IGT), diabetes,or other components of the metabolic syndromefollowing completion of therapy. Establishing thisassociation may have implications for how we followthis subpopulation of patients.
Methods
Subjects
This study was carried out at The Hospital for SickChildren and Princess Margaret Hospital in Toronto,and received approval by the institutional ethics boardof both hospitals. Survivors of pediatric pre B cell andT cell ALL, ages 10 yr and older, who were diagnosed atthe Hospital for Sick Children, between January 1996and December 2006, and had completed treatment atleast 6 months prior to the study were approached toparticipate in the study. This date range was chosento allow for inclusion of subjects at least 6 monthsfrom end of therapy. Medical charts were reviewedand patients who met MID criteria were recruited tothe control group while age matched patients withno MID were recruited to the control group in a 2:1
ratio. The following protocols were used during thestudy period for patients with B lineage ALL: from1998 to 1999, either POG 9605, or our institutional 3drug (protocol AB) or 4 drug (protocol C) regimen;from 2000 to 2004, the POG 9900 series which includesa 3 drug and a 4 drug induction; and from 2004onward, COG AALL0331 or AALL0232. Patientswith T-cell ALL from 1997 to 1999 received protocolC and COG A5971 after 1999. The 3 drug inductionregimen consisted of vincristine, l-asparaginase andeither prednisone or dexamethasone; the 4 drugregimen consisted of vincristine, l-asparaginase,daunomycin, and either prednisone (40 mg/m2/day) ordexamethasone (6 mg/m2/day). Exclusion criteria wereage younger than 12 months of age at the time oftherapy, patients who had a relapse or underwent bonemarrow transplantation, patients with pre-existingdiabetes, or those with testicular involvement atthe time of diagnosis – a finding which significantlychanges the initial protocol. The clinical component ofthe study was conducted at the clinical investigationalunit (CIU) at the Hospital for Sick Children. Consentwas obtained at that visit.
Definition of MID
Hyperglycemia in this population has been defined ina number of different ways in the literature (3, 6, 8, 13,14). In order to isolate a group with more persistenthyperglycemia, only patients with a plasma glucoselevel ≥11.1 mmol/L (200 mg/dL) on two separate daysor more during remission induction, were consideredto have MID. The definition is in accordance with theAmerican Diabetes Association definition (15) and hasbeen used in previous studies of MID in this population(3, 4).
Data collection
Data collection obtained from questionnaire andmedical chart included family history of type 2 diabetes,obesity, cardiovascular disease and dyslipidemia,ethnic background, concurrent illnesses, and use ofmedications. Data pertaining to the illness, includingtreatment protocols, type of glucocorticoid used, andglucose levels during induction was also collected.
Anthropometric measurements including height,weight, waist circumference, and body mass index(BMI) were obtained on the study visit. Waistcircumference to height ratio was calculated and aWC/Ht ratio ≥0.5 was considered increased risk forcentral adiposity and insulin resistance as describedpreviously (16). BMI z-score was calculated usingWHO/WIC growth charts (17), blood pressure wasmeasured for each subject and pubertal stage wasassessed using Tanner staging.
2 Pediatric Diabetes 2014
MID during ALL treatment and metabolic risk
Fasting, total cholesterol, low-density lipoprotein(LDL) cholesterol, high-density lipoprotein (HDL)cholesterol, triglycerides, and alanine amino trans-ferase (ALT) were measured. An oral glucose tolerancetest (OGTT) using a glucose dose of 1.75 g/kg up toa maximum of 75 g, with concurrent measurementsof glucose and insulin every 30 min for 120 min wasperformed.
Data analysis
IGT and diabetes mellitus (DM) were diagnosedaccording to the definition of the American DiabetesAssociation (15). Insulin sensitivity was calculatedusing the dynamic Matsuda whole body insulinsensitivity index (ISogtt) (18), and area-under-the-curve (AUC) glucose and insulin were calculated usingthe trapezoidal rule (19, 20).
In addition to the presence of IGT/DM, othercomponents of the metabolic syndrome wereassessed including triglyceride level above 1.7 mmol/L,HDL < 1.03 mmol/L, systolic blood pressure above the95th percentile for age and gender, and waist circumfer-ence above the 90th percentile for age and gender. Themetabolic syndrome was defined using pediatric criteriafrom the International Diabetes Federation (21).
Statistical analysis
We used frequency distributions and measuresof central tendency. To examine differences ofcharacteristics between the control and study group,a chi-squared and t-test/Wilcoxon statistics wereperformed, depending on the distribution of the data.Logistic regression was also conducted to comparegroup differences while controlling for age.
Results
A total of 90 patients were recruited for the study,30 subjects in the study group and with 60 subjectsacting as controls. Baseline characteristics at thetime of testing are shown in Table 1. The controlgroup was on average further away from the timeof diagnosis with ALL compared with the studygroup (9.2 ± 2.2 vs. 7.9 ± 2.7 yr). The likelihood ofbeing from a non-Caucasian ethnic group was almosttwice as high in the study group compared with thecontrol group. A waist to height ratio of ≥0.5 wasseen in a significantly higher rate in the study groupcompared with the control group (p = 0.01), and therewas a trend toward increased frequency of IGT in thestudy group compared with the control group (4/30 vs.1/90, respectively, p = 0.07). No patients had impairedfasting glucose (IFG) or DM. Using the IDF definitionof metabolic syndrome for this age group (21) we
observed a trend toward higher proportion of patientswith two or more features of the metabolic syndromein the study group (p = 0.09). No significant differencesin lipid levels, insulin secretion, or sensitivity indiceswere noted between the groups (Table 1).
Discussion
MID in children commonly occurs during theinduction phase of treatment for leukemia. Previousstudies have looked at the association between MIDduring induction and the outcome from an oncologyperspective. Studies in adults have shown increasedhospital mortality, shorter remission duration, shortersurvival rates, and higher rates of sepsis in patients whodeveloped MID (22, 23). Similar findings were notedin a single pediatric cohort showing increased infectionrates and poorer survival rates (7, 8), but others havefound no difference in remission, survival, or relapserates between children who did or did not developMID (6). Despite reports regarding an increased riskof metabolic disturbances in survivors of ALL (11, 24),no previous study has assessed whether this metabolicrisk was related to a previous history of MID duringinduction therapy. In this study we hypothesize thatdeveloping MID during induction treatment may bean early surrogate marker for metabolic disturbanceslater in life. Waist-to-height ratio equal or greater than0.5 has previously been shown to be associated withhigher metabolic risk (16, 25–28), and in this study itwas present in higher rates in children with history ofMID compared with children with no history of MID.Rates of IGT were four fold higher in the study groupcompared with the control group although this findingdid not reach statistical significance. This trend seenmay prove to be significant in a larger study or after agreater length of time since therapy in these subjects.
We observed a relatively high frequency of IGTin adolescents with a previous history of MID, mostof who are not considered overweight based upontheir BMI. Therefore, initiation of closer monitoringfor metabolic diseases in this group of patients maybe warranted when they reach young adulthood.Past history of MID may be an important riskfactor in determining stratification of risk for futurecardiometabolic dysfunction. However, our anecdotalexperience suggests that parents do not often knowwhether their child developed or was treated for MIDduring induction and therefore may not be able toprovide this information. Furthermore, MID usuallylasts for a brief time and frequently does not requiretreatment with insulin so may not even be brought toparents attention.
This study raises the question of whether MIDoccurrence during induction treatment for ALLidentifies children at future metabolic risk. During
Pediatric Diabetes 2014 3
Yeshayahu et al.
Table 1. Anthropometric and clinical data of both groups
ALT, alanine amino transferase; AUC, area under the curve; BMI, body mass index; HDL, high-density lipoprotein; IGT,impaired glucose tolerance; ISogtt, Matsuda whole body insulin sensitivity index; LDL, low-density lipoprotein.Parameters which were found to be statistically significant between groups are given in bold.*Percentage of subjects with family history of any of the following: Diabetes, coronary heart disease, hypertension, obesity, ordyslipidemia.
induction therapy, medications that induce insulinresistance or reduce insulin secretion may stress the‘vulnerable ß-cell’ and unmask the pancreas, andtherefore the patient, at future risk for metabolicdysfunction. This may be particularly true if additionalrisk factors are present (e.g., non-Caucasian ethnicgroup, older age at time of induction, or development ofincreased central adiposity as measured by waist/heightratio) all of which were more common in patients whoexperienced MID in our study.
Limitations of this study include a relatively lowfrequency of MID leading to some challenges inrecruiting large numbers of subjects to the study group.Age at the time of assessment was slightly older inthe MID group, however, we adjusted for age in allour analyses to minimize the effect of this differencebetween groups on the results. Baseline informationand differences between the comparison groups inabdominal obesity rates was not available. As thiselement is known to influence metabolic outcomes, thelack of these data may have affected the measuredoutcome.
In conclusion, children with ALL who develop MIDduring the induction treatment may have an increasedrisk for metabolic syndrome features and IGT yearslater.
This is the first study to compare metabolic outcomesof children who developed MID to those whoremained normoglycemic during induction treatment
for leukemia. Our study demonstrates a tendencytoward increased risk for IGT and multiple metabolicsyndrome features in children with a history of MID.Further research is needed to confirm this findingalong with longitudinal studies to determine if riskincreases more with age. This may indicate the need forongoing metabolic screening and institution of lifestylemodifications to prevent future disease.
Acknowledgements
The authors wish to thank Dr David Hodgson from PrincessMargaret Hospital in Toronto for his assistance in recruitingpatients, and to Ms Junlang Yin for the statistical analysis.
References
1. Howlader NNA, Krapcho M, Neyman N, Aminou
R, Waldron W. SEER Cancer Statistics Review,1975–2008. Bethesda, MD: National Cancer Institute,2011.
2. Pui CH, Burghen GA, Bowman WP, Aur RJ. Riskfactors for hyperglycemia in children with leukemiareceiving L-asparaginase and prednisone. J Pediatr1981: 99: 46–50.
3. Lowas SR, Marks D, Malempati S. Prevalence of tran-sient hyperglycemia during induction chemotherapy forpediatric acute lymphoblastic leukemia. Pediatr BloodCancer 2009: 52: 814–818.
4. Baillargeon J, Langevin AM, Mullins J et al.Transient hyperglycemia in Hispanic children with acute
4 Pediatric Diabetes 2014
MID during ALL treatment and metabolic risk
lymphoblastic leukemia. Pediatr Blood Cancer 2005:45: 960–963.
5. Koltin D, Sung L, Naqvi A, Urbach SL. Medicationinduced diabetes during induction in pediatric acutelymphoblastic leukemia: prevalence, risk factors andcharacteristics. Support Care Cancer 2012: 20:2009–2015.
6. Roberson JR, Spraker HL, Shelso J et al. Clinicalconsequences of hyperglycemia during remissioninduction therapy for pediatric acute lymphoblasticleukemia. Leukemia 2009: 23: 245–250.
7. Sonabend RY, McKay SV, Okcu MF, Yan J,Haymond MW, Margolin JF. Hyperglycemia duringinduction therapy is associated with increased infectiouscomplications in childhood acute lymphocytic leukemia.Pediatr Blood Cancer 2008: 51: 387–392.
8. Sonabend RY, McKay SV, Okcu MF, Yan J,Haymond MW, Margolin JF. Hyperglycemia duringinduction therapy is associated with poorer survival inchildren with acute lymphocytic leukemia. J Pediatr2009: 155: 73–78.
9. Pui CH, Pei D, Campana D et al. Improved prognosisfor older adolescents with acute lymphoblasticleukemia. J Clin Oncol 2011: 29: 386–391.
10. Mertens AC, Yasui Y, Neglia JP et al. Late mortalityexperience in five-year survivors of childhood andadolescent cancer: the Childhood Cancer SurvivorStudy. J Clin Oncol 2001: 19: 3163–3172.
11. Oudin C, Simeoni MC, Sirvent N et al. Prevalence andrisk factors of the metabolic syndrome in adult survivorsof childhood leukemia. Blood 2011: 117: 4442–4448.
12. Surapolchai P, Hongeng S, Mahachoklertwat-
tana P et al. Impaired glucose tolerance andinsulin resistance in survivors of childhood acutelymphoblastic leukemia: prevalence and risk fac-tors. J Pediatr Hematol Oncol 2010: 32: 383–389.doi:10.1097/MPH.0b013e3181dccc0b.
13. Pui CH, Burghen GA, Bowman WP, Aur RJ. Riskfactors for hyperglycemia in children with leukemiareceiving L-asparaginase and prednisone. J Pediatr1981: 99: 46–50.
14. Baillargeon J, Langevin AM, Mullins J et al.Transient hyperglycemia in Hispanic children with acutelymphoblastic leukemia. Pediatr Blood Cancer 2005: 45:960–963.
15. American Diabetes Association. Diagnosis and classi-fication of diabetes mellitus. Diabetes Care 2006: 29(Suppl. 1): S43–S48.
16. Garnett SP, Baur LA, Cowell CT. Waist-to-heightratio: a simple option for determining excess centraladiposity in young people. Int J Obes (Lond) 2008: 32:1028–1030.
17. WHO Multicentre Growth Reference Study Group.WHO Child Growth Standards based on length/height,weight and age. Acta Paediatr 2006: 450: 76–85.
18. Matsuda M, DeFronzo RA. Insulin sensitivityindices obtained from oral glucose tolerance testing:comparison with the euglycemic insulin clamp. DiabetesCare 1999: 22: 1462–1470.
19. Ahren B, Pacini G. Importance of quantifying insulinsecretion in relation to insulin sensitivity to accuratelyassess beta cell function in clinical studies. Eur JEndocrinol 2004: 150: 97–104.
20. Phillips DI, Clark PM, Hales CN, Osmond C.Understanding oral glucose tolerance: comparison ofglucose or insulin measurements during the oral glucosetolerance test with specific measurements of insulinresistance and insulin secretion. Diabet Med 1994: 11:286–292.
21. Zimmet P, Alberti KG, Kaufman F et al. Themetabolic syndrome in children and adolescents -an IDF consensus report. Pediatr Diabetes 2007: 8:299–306.
22. Ali NA, O’Brien JM Jr, Blum W et al. Hyperglycemiain patients with acute myeloid leukemia is associatedwith increased hospital mortality. Cancer 2007: 110:96–102.
23. Weiser MA, Cabanillas ME, Konopleva M et al.Relation between the duration of remission andhyperglycemia during induction chemotherapy foracute lymphocytic leukemia with a hyperfraction-ated cyclophosphamide, vincristine, doxorubicin, anddexamethasone/methotrexate-cytarabine regimen. Can-cer 2004: 100: 1179–1185.
24. Mohn A, Di Marzio A, Capanna R, Fioritoni G,Chiarelli F. Persistence of impaired pancreatic beta-cell function in children treated for acute lymphoblasticleukaemia. Lancet 2004: 363: 127–128.
25. Goulding A, Taylor RW, Grant AM, Parnell
WR, Wilson NC, Williams SM. Waist-to-heightratios in relation to BMI z-scores in three ethnicgroups from a representative sample of New Zealandchildren aged 5–14 years. Int J Obes (Lond) 2010: 34:1188–1190.
26. Savva SC, Tornaritis M, Savva ME et al. Waistcircumference and waist-to-height ratio are betterpredictors of cardiovascular disease risk factors inchildren than body mass index. Int J Obes Relat MetabDisord 2000: 24: 1453–1458.
27. Taylor RW, Brooking L, Williams SM et al. Bodymass index and waist circumference cutoffs to defineobesity in indigenous New Zealanders. Am J Clin Nutr2010: 92: 390–397.
28. Taylor RW, Williams SM, Grant AM, Taylor BJ,Goulding A. Predictive ability of waist-to-height inrelation to adiposity in children is not improved withage and sex-specific values. Obesity (Silver Spring) 2011:19: 1062–1068.
![MEDICATION THERAPY - moh.gov. 2010] Protocol Medication... · Medication Therapy Adherence Clinic (MTAC)…](https://static.documents.pub/doc/80x56/5ca02a4e88c993c0178cbd41/medication-therapy-mohgov-2010-protocol-medication-medication-therapy.jpg)
